Fathom: upload expert-e1-static/training_log.json
Browse files
adapters/expert-e1-static/training_log.json
ADDED
|
@@ -0,0 +1,1141 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"loss": 1.2771,
|
| 4 |
+
"grad_norm": 0.48179489374160767,
|
| 5 |
+
"learning_rate": 9e-06,
|
| 6 |
+
"entropy": 1.0744440883398056,
|
| 7 |
+
"num_tokens": 384772.0,
|
| 8 |
+
"mean_token_accuracy": 0.7264965653419495,
|
| 9 |
+
"epoch": 0.008849557522123894,
|
| 10 |
+
"step": 10
|
| 11 |
+
},
|
| 12 |
+
{
|
| 13 |
+
"loss": 1.2083,
|
| 14 |
+
"grad_norm": 0.22539448738098145,
|
| 15 |
+
"learning_rate": 1.9e-05,
|
| 16 |
+
"entropy": 1.0771890074014663,
|
| 17 |
+
"num_tokens": 794683.0,
|
| 18 |
+
"mean_token_accuracy": 0.7380238056182862,
|
| 19 |
+
"epoch": 0.017699115044247787,
|
| 20 |
+
"step": 20
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"loss": 1.1201,
|
| 24 |
+
"grad_norm": 0.22262926399707794,
|
| 25 |
+
"learning_rate": 2.9e-05,
|
| 26 |
+
"entropy": 1.1027102798223496,
|
| 27 |
+
"num_tokens": 1216610.0,
|
| 28 |
+
"mean_token_accuracy": 0.7497482478618622,
|
| 29 |
+
"epoch": 0.02654867256637168,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"loss": 1.0997,
|
| 34 |
+
"grad_norm": 0.2455902248620987,
|
| 35 |
+
"learning_rate": 3.9000000000000006e-05,
|
| 36 |
+
"entropy": 1.13816859126091,
|
| 37 |
+
"num_tokens": 1636160.0,
|
| 38 |
+
"mean_token_accuracy": 0.751062735915184,
|
| 39 |
+
"epoch": 0.035398230088495575,
|
| 40 |
+
"step": 40
|
| 41 |
+
},
|
| 42 |
+
{
|
| 43 |
+
"loss": 0.9732,
|
| 44 |
+
"grad_norm": 0.25331178307533264,
|
| 45 |
+
"learning_rate": 4.9e-05,
|
| 46 |
+
"entropy": 0.967724832892418,
|
| 47 |
+
"num_tokens": 2047660.0,
|
| 48 |
+
"mean_token_accuracy": 0.7753094494342804,
|
| 49 |
+
"epoch": 0.04424778761061947,
|
| 50 |
+
"step": 50
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"loss": 0.8376,
|
| 54 |
+
"grad_norm": 0.2812139391899109,
|
| 55 |
+
"learning_rate": 5.9e-05,
|
| 56 |
+
"entropy": 0.8342341184616089,
|
| 57 |
+
"num_tokens": 2442064.0,
|
| 58 |
+
"mean_token_accuracy": 0.8068537533283233,
|
| 59 |
+
"epoch": 0.05309734513274336,
|
| 60 |
+
"step": 60
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"loss": 0.7207,
|
| 64 |
+
"grad_norm": 0.5564813613891602,
|
| 65 |
+
"learning_rate": 6.9e-05,
|
| 66 |
+
"entropy": 0.7252395421266555,
|
| 67 |
+
"num_tokens": 2867532.0,
|
| 68 |
+
"mean_token_accuracy": 0.832821500301361,
|
| 69 |
+
"epoch": 0.061946902654867256,
|
| 70 |
+
"step": 70
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"loss": 0.6873,
|
| 74 |
+
"grad_norm": 0.28080248832702637,
|
| 75 |
+
"learning_rate": 7.900000000000001e-05,
|
| 76 |
+
"entropy": 0.6883613392710686,
|
| 77 |
+
"num_tokens": 3271081.0,
|
| 78 |
+
"mean_token_accuracy": 0.8418284684419632,
|
| 79 |
+
"epoch": 0.07079646017699115,
|
| 80 |
+
"step": 80
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"loss": 0.6869,
|
| 84 |
+
"grad_norm": 0.37117382884025574,
|
| 85 |
+
"learning_rate": 8.900000000000001e-05,
|
| 86 |
+
"entropy": 0.6866619646549225,
|
| 87 |
+
"num_tokens": 3676002.0,
|
| 88 |
+
"mean_token_accuracy": 0.8389175772666931,
|
| 89 |
+
"epoch": 0.07964601769911504,
|
| 90 |
+
"step": 90
|
| 91 |
+
},
|
| 92 |
+
{
|
| 93 |
+
"loss": 0.641,
|
| 94 |
+
"grad_norm": 0.34726274013519287,
|
| 95 |
+
"learning_rate": 9.900000000000001e-05,
|
| 96 |
+
"entropy": 0.6441589877009392,
|
| 97 |
+
"num_tokens": 4109097.0,
|
| 98 |
+
"mean_token_accuracy": 0.849273630976677,
|
| 99 |
+
"epoch": 0.08849557522123894,
|
| 100 |
+
"step": 100
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"loss": 0.541,
|
| 104 |
+
"grad_norm": 0.3802001476287842,
|
| 105 |
+
"learning_rate": 9.998116250927091e-05,
|
| 106 |
+
"entropy": 0.54506506472826,
|
| 107 |
+
"num_tokens": 4518063.0,
|
| 108 |
+
"mean_token_accuracy": 0.8707349866628646,
|
| 109 |
+
"epoch": 0.09734513274336283,
|
| 110 |
+
"step": 110
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"loss": 0.5358,
|
| 114 |
+
"grad_norm": 0.33883532881736755,
|
| 115 |
+
"learning_rate": 9.991606348016586e-05,
|
| 116 |
+
"entropy": 0.5439137145876884,
|
| 117 |
+
"num_tokens": 4942538.0,
|
| 118 |
+
"mean_token_accuracy": 0.8710391223430634,
|
| 119 |
+
"epoch": 0.10619469026548672,
|
| 120 |
+
"step": 120
|
| 121 |
+
},
|
| 122 |
+
{
|
| 123 |
+
"loss": 0.513,
|
| 124 |
+
"grad_norm": 0.2743840515613556,
|
| 125 |
+
"learning_rate": 9.980453089389592e-05,
|
| 126 |
+
"entropy": 0.530657310783863,
|
| 127 |
+
"num_tokens": 5366333.0,
|
| 128 |
+
"mean_token_accuracy": 0.8756012558937073,
|
| 129 |
+
"epoch": 0.11504424778761062,
|
| 130 |
+
"step": 130
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"loss": 0.499,
|
| 134 |
+
"grad_norm": 0.3075107932090759,
|
| 135 |
+
"learning_rate": 9.964666850172589e-05,
|
| 136 |
+
"entropy": 0.5022920548915863,
|
| 137 |
+
"num_tokens": 5770091.0,
|
| 138 |
+
"mean_token_accuracy": 0.8816363781690597,
|
| 139 |
+
"epoch": 0.12389380530973451,
|
| 140 |
+
"step": 140
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"loss": 0.4638,
|
| 144 |
+
"grad_norm": 0.3095848560333252,
|
| 145 |
+
"learning_rate": 9.944262315242346e-05,
|
| 146 |
+
"entropy": 0.47086685746908186,
|
| 147 |
+
"num_tokens": 6186047.0,
|
| 148 |
+
"mean_token_accuracy": 0.8876948118209839,
|
| 149 |
+
"epoch": 0.13274336283185842,
|
| 150 |
+
"step": 150
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"loss": 0.4325,
|
| 154 |
+
"grad_norm": 0.28685304522514343,
|
| 155 |
+
"learning_rate": 9.919258465565575e-05,
|
| 156 |
+
"entropy": 0.43743315935134885,
|
| 157 |
+
"num_tokens": 6585548.0,
|
| 158 |
+
"mean_token_accuracy": 0.8952628016471863,
|
| 159 |
+
"epoch": 0.1415929203539823,
|
| 160 |
+
"step": 160
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"loss": 0.4701,
|
| 164 |
+
"grad_norm": 0.324359267950058,
|
| 165 |
+
"learning_rate": 9.889678560542202e-05,
|
| 166 |
+
"entropy": 0.4746619015932083,
|
| 167 |
+
"num_tokens": 7013633.0,
|
| 168 |
+
"mean_token_accuracy": 0.8870637893676758,
|
| 169 |
+
"epoch": 0.1504424778761062,
|
| 170 |
+
"step": 170
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"loss": 0.4442,
|
| 174 |
+
"grad_norm": 0.31656569242477417,
|
| 175 |
+
"learning_rate": 9.855550116368716e-05,
|
| 176 |
+
"entropy": 0.4526193618774414,
|
| 177 |
+
"num_tokens": 7419505.0,
|
| 178 |
+
"mean_token_accuracy": 0.8906016558408737,
|
| 179 |
+
"epoch": 0.1592920353982301,
|
| 180 |
+
"step": 180
|
| 181 |
+
},
|
| 182 |
+
{
|
| 183 |
+
"loss": 0.412,
|
| 184 |
+
"grad_norm": 0.30724215507507324,
|
| 185 |
+
"learning_rate": 9.816904880441713e-05,
|
| 186 |
+
"entropy": 0.4151700422167778,
|
| 187 |
+
"num_tokens": 7818357.0,
|
| 188 |
+
"mean_token_accuracy": 0.8987403184175491,
|
| 189 |
+
"epoch": 0.168141592920354,
|
| 190 |
+
"step": 190
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"loss": 0.4121,
|
| 194 |
+
"grad_norm": 0.49001774191856384,
|
| 195 |
+
"learning_rate": 9.773778801825405e-05,
|
| 196 |
+
"entropy": 0.4183415174484253,
|
| 197 |
+
"num_tokens": 8275452.0,
|
| 198 |
+
"mean_token_accuracy": 0.898220694065094,
|
| 199 |
+
"epoch": 0.17699115044247787,
|
| 200 |
+
"step": 200
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"loss": 0.4359,
|
| 204 |
+
"grad_norm": 0.5563516616821289,
|
| 205 |
+
"learning_rate": 9.726211997810646e-05,
|
| 206 |
+
"entropy": 0.44024592265486717,
|
| 207 |
+
"num_tokens": 8689265.0,
|
| 208 |
+
"mean_token_accuracy": 0.8926420778036117,
|
| 209 |
+
"epoch": 0.18584070796460178,
|
| 210 |
+
"step": 210
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"loss": 0.4265,
|
| 214 |
+
"grad_norm": 0.4248098134994507,
|
| 215 |
+
"learning_rate": 9.674248716596497e-05,
|
| 216 |
+
"entropy": 0.43353245332837104,
|
| 217 |
+
"num_tokens": 9108244.0,
|
| 218 |
+
"mean_token_accuracy": 0.8954088747501373,
|
| 219 |
+
"epoch": 0.19469026548672566,
|
| 220 |
+
"step": 220
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"loss": 0.4113,
|
| 224 |
+
"grad_norm": 0.5417165160179138,
|
| 225 |
+
"learning_rate": 9.617937296129115e-05,
|
| 226 |
+
"entropy": 0.4203181877732277,
|
| 227 |
+
"num_tokens": 9508482.0,
|
| 228 |
+
"mean_token_accuracy": 0.8986817359924316,
|
| 229 |
+
"epoch": 0.20353982300884957,
|
| 230 |
+
"step": 230
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"loss": 0.4041,
|
| 234 |
+
"grad_norm": 0.37166956067085266,
|
| 235 |
+
"learning_rate": 9.557330119136203e-05,
|
| 236 |
+
"entropy": 0.4078876577317715,
|
| 237 |
+
"num_tokens": 9929607.0,
|
| 238 |
+
"mean_token_accuracy": 0.9026990383863449,
|
| 239 |
+
"epoch": 0.21238938053097345,
|
| 240 |
+
"step": 240
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"loss": 0.3372,
|
| 244 |
+
"grad_norm": 0.37398409843444824,
|
| 245 |
+
"learning_rate": 9.492483564398883e-05,
|
| 246 |
+
"entropy": 0.3507886216044426,
|
| 247 |
+
"num_tokens": 10352037.0,
|
| 248 |
+
"mean_token_accuracy": 0.9149532109498978,
|
| 249 |
+
"epoch": 0.22123893805309736,
|
| 250 |
+
"step": 250
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"loss": 0.3754,
|
| 254 |
+
"grad_norm": 0.4071371853351593,
|
| 255 |
+
"learning_rate": 9.423457954306312e-05,
|
| 256 |
+
"entropy": 0.3833847470581532,
|
| 257 |
+
"num_tokens": 10751656.0,
|
| 258 |
+
"mean_token_accuracy": 0.9078819990158081,
|
| 259 |
+
"epoch": 0.23008849557522124,
|
| 260 |
+
"step": 260
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"loss": 0.3092,
|
| 264 |
+
"grad_norm": 0.3898226022720337,
|
| 265 |
+
"learning_rate": 9.350317498741811e-05,
|
| 266 |
+
"entropy": 0.31048981472849846,
|
| 267 |
+
"num_tokens": 11177021.0,
|
| 268 |
+
"mean_token_accuracy": 0.9235329061746598,
|
| 269 |
+
"epoch": 0.23893805309734514,
|
| 270 |
+
"step": 270
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"loss": 0.2992,
|
| 274 |
+
"grad_norm": 0.4743488132953644,
|
| 275 |
+
"learning_rate": 9.273130235352743e-05,
|
| 276 |
+
"entropy": 0.3064242236316204,
|
| 277 |
+
"num_tokens": 11589241.0,
|
| 278 |
+
"mean_token_accuracy": 0.926247027516365,
|
| 279 |
+
"epoch": 0.24778761061946902,
|
| 280 |
+
"step": 280
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"loss": 0.3432,
|
| 284 |
+
"grad_norm": 0.3992835283279419,
|
| 285 |
+
"learning_rate": 9.191967966259645e-05,
|
| 286 |
+
"entropy": 0.35372441038489344,
|
| 287 |
+
"num_tokens": 11997925.0,
|
| 288 |
+
"mean_token_accuracy": 0.914875665307045,
|
| 289 |
+
"epoch": 0.25663716814159293,
|
| 290 |
+
"step": 290
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"loss": 0.311,
|
| 294 |
+
"grad_norm": 0.4893634021282196,
|
| 295 |
+
"learning_rate": 9.10690619126356e-05,
|
| 296 |
+
"entropy": 0.31616691052913665,
|
| 297 |
+
"num_tokens": 12415929.0,
|
| 298 |
+
"mean_token_accuracy": 0.9236188173294068,
|
| 299 |
+
"epoch": 0.26548672566371684,
|
| 300 |
+
"step": 300
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"loss": 0.3358,
|
| 304 |
+
"grad_norm": 0.38468873500823975,
|
| 305 |
+
"learning_rate": 9.018024037613646e-05,
|
| 306 |
+
"entropy": 0.3419831946492195,
|
| 307 |
+
"num_tokens": 12828979.0,
|
| 308 |
+
"mean_token_accuracy": 0.9184246301651001,
|
| 309 |
+
"epoch": 0.2743362831858407,
|
| 310 |
+
"step": 310
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"loss": 0.3176,
|
| 314 |
+
"grad_norm": 0.4959142208099365,
|
| 315 |
+
"learning_rate": 8.925404186400408e-05,
|
| 316 |
+
"entropy": 0.3222149942070246,
|
| 317 |
+
"num_tokens": 13222249.0,
|
| 318 |
+
"mean_token_accuracy": 0.921812680363655,
|
| 319 |
+
"epoch": 0.2831858407079646,
|
| 320 |
+
"step": 320
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"loss": 0.2962,
|
| 324 |
+
"grad_norm": 0.33697429299354553,
|
| 325 |
+
"learning_rate": 8.829132795643051e-05,
|
| 326 |
+
"entropy": 0.30270505994558333,
|
| 327 |
+
"num_tokens": 13653508.0,
|
| 328 |
+
"mean_token_accuracy": 0.9271777629852295,
|
| 329 |
+
"epoch": 0.2920353982300885,
|
| 330 |
+
"step": 330
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"loss": 0.2721,
|
| 334 |
+
"grad_norm": 0.3220033645629883,
|
| 335 |
+
"learning_rate": 8.729299420142465e-05,
|
| 336 |
+
"entropy": 0.2873320683836937,
|
| 337 |
+
"num_tokens": 14076074.0,
|
| 338 |
+
"mean_token_accuracy": 0.9307485371828079,
|
| 339 |
+
"epoch": 0.3008849557522124,
|
| 340 |
+
"step": 340
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"loss": 0.3197,
|
| 344 |
+
"grad_norm": 0.3136242628097534,
|
| 345 |
+
"learning_rate": 8.625996928174412e-05,
|
| 346 |
+
"entropy": 0.332285375893116,
|
| 347 |
+
"num_tokens": 14496295.0,
|
| 348 |
+
"mean_token_accuracy": 0.9201685935258865,
|
| 349 |
+
"epoch": 0.30973451327433627,
|
| 350 |
+
"step": 350
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"loss": 0.3053,
|
| 354 |
+
"grad_norm": 0.43096667528152466,
|
| 355 |
+
"learning_rate": 8.519321415100414e-05,
|
| 356 |
+
"entropy": 0.310824865847826,
|
| 357 |
+
"num_tokens": 14897719.0,
|
| 358 |
+
"mean_token_accuracy": 0.924813660979271,
|
| 359 |
+
"epoch": 0.3185840707964602,
|
| 360 |
+
"step": 360
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"loss": 0.2486,
|
| 364 |
+
"grad_norm": 0.4815881848335266,
|
| 365 |
+
"learning_rate": 8.409372113976712e-05,
|
| 366 |
+
"entropy": 0.2627465195953846,
|
| 367 |
+
"num_tokens": 15293602.0,
|
| 368 |
+
"mean_token_accuracy": 0.9371735215187073,
|
| 369 |
+
"epoch": 0.3274336283185841,
|
| 370 |
+
"step": 370
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"loss": 0.2657,
|
| 374 |
+
"grad_norm": 0.4008077085018158,
|
| 375 |
+
"learning_rate": 8.296251303244413e-05,
|
| 376 |
+
"entropy": 0.26940090730786326,
|
| 377 |
+
"num_tokens": 15726200.0,
|
| 378 |
+
"mean_token_accuracy": 0.9358527153730393,
|
| 379 |
+
"epoch": 0.336283185840708,
|
| 380 |
+
"step": 380
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"loss": 0.2387,
|
| 384 |
+
"grad_norm": 0.3524218201637268,
|
| 385 |
+
"learning_rate": 8.180064211586738e-05,
|
| 386 |
+
"entropy": 0.24938426464796065,
|
| 387 |
+
"num_tokens": 16150169.0,
|
| 388 |
+
"mean_token_accuracy": 0.9398461669683457,
|
| 389 |
+
"epoch": 0.34513274336283184,
|
| 390 |
+
"step": 390
|
| 391 |
+
},
|
| 392 |
+
{
|
| 393 |
+
"loss": 0.2634,
|
| 394 |
+
"grad_norm": 0.35844066739082336,
|
| 395 |
+
"learning_rate": 8.060918920041856e-05,
|
| 396 |
+
"entropy": 0.2718000315129757,
|
| 397 |
+
"num_tokens": 16558293.0,
|
| 398 |
+
"mean_token_accuracy": 0.9348433256149292,
|
| 399 |
+
"epoch": 0.35398230088495575,
|
| 400 |
+
"step": 400
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"loss": 0.2835,
|
| 404 |
+
"grad_norm": 0.46172916889190674,
|
| 405 |
+
"learning_rate": 7.938926261462366e-05,
|
| 406 |
+
"entropy": 0.2921802319586277,
|
| 407 |
+
"num_tokens": 16997838.0,
|
| 408 |
+
"mean_token_accuracy": 0.9297347247600556,
|
| 409 |
+
"epoch": 0.36283185840707965,
|
| 410 |
+
"step": 410
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"loss": 0.2743,
|
| 414 |
+
"grad_norm": 0.3828360140323639,
|
| 415 |
+
"learning_rate": 7.81419971741494e-05,
|
| 416 |
+
"entropy": 0.2844072911888361,
|
| 417 |
+
"num_tokens": 17414034.0,
|
| 418 |
+
"mean_token_accuracy": 0.9316927522420884,
|
| 419 |
+
"epoch": 0.37168141592920356,
|
| 420 |
+
"step": 420
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"loss": 0.2481,
|
| 424 |
+
"grad_norm": 0.2896506190299988,
|
| 425 |
+
"learning_rate": 7.686855312616055e-05,
|
| 426 |
+
"entropy": 0.2565284200012684,
|
| 427 |
+
"num_tokens": 17851569.0,
|
| 428 |
+
"mean_token_accuracy": 0.9383330553770065,
|
| 429 |
+
"epoch": 0.3805309734513274,
|
| 430 |
+
"step": 430
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"loss": 0.278,
|
| 434 |
+
"grad_norm": 0.5132155418395996,
|
| 435 |
+
"learning_rate": 7.557011507002004e-05,
|
| 436 |
+
"entropy": 0.2847262255847454,
|
| 437 |
+
"num_tokens": 18265230.0,
|
| 438 |
+
"mean_token_accuracy": 0.9326902598142623,
|
| 439 |
+
"epoch": 0.3893805309734513,
|
| 440 |
+
"step": 440
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"loss": 0.2172,
|
| 444 |
+
"grad_norm": 0.36239564418792725,
|
| 445 |
+
"learning_rate": 7.424789085533584e-05,
|
| 446 |
+
"entropy": 0.2237854577600956,
|
| 447 |
+
"num_tokens": 18669967.0,
|
| 448 |
+
"mean_token_accuracy": 0.9465960383415222,
|
| 449 |
+
"epoch": 0.39823008849557523,
|
| 450 |
+
"step": 450
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"loss": 0.2868,
|
| 454 |
+
"grad_norm": 0.263062059879303,
|
| 455 |
+
"learning_rate": 7.290311045837963e-05,
|
| 456 |
+
"entropy": 0.29681268632411956,
|
| 457 |
+
"num_tokens": 19100082.0,
|
| 458 |
+
"mean_token_accuracy": 0.9297707736492157,
|
| 459 |
+
"epoch": 0.40707964601769914,
|
| 460 |
+
"step": 460
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"loss": 0.2535,
|
| 464 |
+
"grad_norm": 0.25322869420051575,
|
| 465 |
+
"learning_rate": 7.153702483792266e-05,
|
| 466 |
+
"entropy": 0.2623436853289604,
|
| 467 |
+
"num_tokens": 19535145.0,
|
| 468 |
+
"mean_token_accuracy": 0.9386782139539719,
|
| 469 |
+
"epoch": 0.415929203539823,
|
| 470 |
+
"step": 470
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"loss": 0.2781,
|
| 474 |
+
"grad_norm": 0.32808470726013184,
|
| 475 |
+
"learning_rate": 7.015090477155288e-05,
|
| 476 |
+
"entropy": 0.2793763652443886,
|
| 477 |
+
"num_tokens": 19946800.0,
|
| 478 |
+
"mean_token_accuracy": 0.9325116276741028,
|
| 479 |
+
"epoch": 0.4247787610619469,
|
| 480 |
+
"step": 480
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"loss": 0.2166,
|
| 484 |
+
"grad_norm": 0.3056963086128235,
|
| 485 |
+
"learning_rate": 6.874603967355603e-05,
|
| 486 |
+
"entropy": 0.2249807395040989,
|
| 487 |
+
"num_tokens": 20374389.0,
|
| 488 |
+
"mean_token_accuracy": 0.9458037465810776,
|
| 489 |
+
"epoch": 0.4336283185840708,
|
| 490 |
+
"step": 490
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"loss": 0.2528,
|
| 494 |
+
"grad_norm": 0.3542698919773102,
|
| 495 |
+
"learning_rate": 6.732373639546025e-05,
|
| 496 |
+
"entropy": 0.2597987335175276,
|
| 497 |
+
"num_tokens": 20792226.0,
|
| 498 |
+
"mean_token_accuracy": 0.9383546471595764,
|
| 499 |
+
"epoch": 0.4424778761061947,
|
| 500 |
+
"step": 500
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"loss": 0.2007,
|
| 504 |
+
"grad_norm": 0.2978123426437378,
|
| 505 |
+
"learning_rate": 6.588531801035993e-05,
|
| 506 |
+
"entropy": 0.2049595110118389,
|
| 507 |
+
"num_tokens": 21206944.0,
|
| 508 |
+
"mean_token_accuracy": 0.9496358513832093,
|
| 509 |
+
"epoch": 0.45132743362831856,
|
| 510 |
+
"step": 510
|
| 511 |
+
},
|
| 512 |
+
{
|
| 513 |
+
"loss": 0.2147,
|
| 514 |
+
"grad_norm": 0.3151644766330719,
|
| 515 |
+
"learning_rate": 6.443212258214983e-05,
|
| 516 |
+
"entropy": 0.221473628282547,
|
| 517 |
+
"num_tokens": 21597252.0,
|
| 518 |
+
"mean_token_accuracy": 0.9471895158290863,
|
| 519 |
+
"epoch": 0.46017699115044247,
|
| 520 |
+
"step": 520
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"loss": 0.2534,
|
| 524 |
+
"grad_norm": 0.28072014451026917,
|
| 525 |
+
"learning_rate": 6.296550192081421e-05,
|
| 526 |
+
"entropy": 0.25930051133036613,
|
| 527 |
+
"num_tokens": 22019525.0,
|
| 528 |
+
"mean_token_accuracy": 0.9380945324897766,
|
| 529 |
+
"epoch": 0.4690265486725664,
|
| 530 |
+
"step": 530
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"loss": 0.2773,
|
| 534 |
+
"grad_norm": 0.28228074312210083,
|
| 535 |
+
"learning_rate": 6.148682032492894e-05,
|
| 536 |
+
"entropy": 0.2860624067485332,
|
| 537 |
+
"num_tokens": 22450805.0,
|
| 538 |
+
"mean_token_accuracy": 0.9322474598884583,
|
| 539 |
+
"epoch": 0.4778761061946903,
|
| 540 |
+
"step": 540
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"loss": 0.2331,
|
| 544 |
+
"grad_norm": 0.22247536480426788,
|
| 545 |
+
"learning_rate": 5.999745331254616e-05,
|
| 546 |
+
"entropy": 0.2394201297312975,
|
| 547 |
+
"num_tokens": 22860630.0,
|
| 548 |
+
"mean_token_accuracy": 0.9421512335538864,
|
| 549 |
+
"epoch": 0.48672566371681414,
|
| 550 |
+
"step": 550
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"loss": 0.2545,
|
| 554 |
+
"grad_norm": 0.36099445819854736,
|
| 555 |
+
"learning_rate": 5.849878634164251e-05,
|
| 556 |
+
"entropy": 0.26039876230061054,
|
| 557 |
+
"num_tokens": 23276454.0,
|
| 558 |
+
"mean_token_accuracy": 0.9390616893768311,
|
| 559 |
+
"epoch": 0.49557522123893805,
|
| 560 |
+
"step": 560
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"loss": 0.2566,
|
| 564 |
+
"grad_norm": 0.24887335300445557,
|
| 565 |
+
"learning_rate": 5.699221352132059e-05,
|
| 566 |
+
"entropy": 0.2648449897766113,
|
| 567 |
+
"num_tokens": 23710582.0,
|
| 568 |
+
"mean_token_accuracy": 0.9375686824321747,
|
| 569 |
+
"epoch": 0.504424778761062,
|
| 570 |
+
"step": 570
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"loss": 0.2481,
|
| 574 |
+
"grad_norm": 0.26001015305519104,
|
| 575 |
+
"learning_rate": 5.547913631496306e-05,
|
| 576 |
+
"entropy": 0.2597955591976643,
|
| 577 |
+
"num_tokens": 24130902.0,
|
| 578 |
+
"mean_token_accuracy": 0.9392797619104385,
|
| 579 |
+
"epoch": 0.5132743362831859,
|
| 580 |
+
"step": 580
|
| 581 |
+
},
|
| 582 |
+
{
|
| 583 |
+
"loss": 0.2491,
|
| 584 |
+
"grad_norm": 0.25495871901512146,
|
| 585 |
+
"learning_rate": 5.396096223654561e-05,
|
| 586 |
+
"entropy": 0.24957055263221264,
|
| 587 |
+
"num_tokens": 24565476.0,
|
| 588 |
+
"mean_token_accuracy": 0.9401059657335281,
|
| 589 |
+
"epoch": 0.5221238938053098,
|
| 590 |
+
"step": 590
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"loss": 0.2479,
|
| 594 |
+
"grad_norm": 0.28261682391166687,
|
| 595 |
+
"learning_rate": 5.2439103541321144e-05,
|
| 596 |
+
"entropy": 0.2580213598906994,
|
| 597 |
+
"num_tokens": 24971253.0,
|
| 598 |
+
"mean_token_accuracy": 0.9390650987625122,
|
| 599 |
+
"epoch": 0.5309734513274337,
|
| 600 |
+
"step": 600
|
| 601 |
+
},
|
| 602 |
+
{
|
| 603 |
+
"loss": 0.2856,
|
| 604 |
+
"grad_norm": 0.2623113691806793,
|
| 605 |
+
"learning_rate": 5.0914975912093854e-05,
|
| 606 |
+
"entropy": 0.2843888055533171,
|
| 607 |
+
"num_tokens": 25375477.0,
|
| 608 |
+
"mean_token_accuracy": 0.9336756229400635,
|
| 609 |
+
"epoch": 0.5398230088495575,
|
| 610 |
+
"step": 610
|
| 611 |
+
},
|
| 612 |
+
{
|
| 613 |
+
"loss": 0.2338,
|
| 614 |
+
"grad_norm": 0.3092866539955139,
|
| 615 |
+
"learning_rate": 4.938999714230467e-05,
|
| 616 |
+
"entropy": 0.23968622721731664,
|
| 617 |
+
"num_tokens": 25767651.0,
|
| 618 |
+
"mean_token_accuracy": 0.9425904780626297,
|
| 619 |
+
"epoch": 0.5486725663716814,
|
| 620 |
+
"step": 620
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"loss": 0.2382,
|
| 624 |
+
"grad_norm": 0.20757952332496643,
|
| 625 |
+
"learning_rate": 4.78655858171533e-05,
|
| 626 |
+
"entropy": 0.24690472409129144,
|
| 627 |
+
"num_tokens": 26197439.0,
|
| 628 |
+
"mean_token_accuracy": 0.9411264300346375,
|
| 629 |
+
"epoch": 0.5575221238938053,
|
| 630 |
+
"step": 630
|
| 631 |
+
},
|
| 632 |
+
{
|
| 633 |
+
"loss": 0.2557,
|
| 634 |
+
"grad_norm": 0.24662645161151886,
|
| 635 |
+
"learning_rate": 4.634315999398393e-05,
|
| 636 |
+
"entropy": 0.2634947098791599,
|
| 637 |
+
"num_tokens": 26612051.0,
|
| 638 |
+
"mean_token_accuracy": 0.9378427803516388,
|
| 639 |
+
"epoch": 0.5663716814159292,
|
| 640 |
+
"step": 640
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"loss": 0.1985,
|
| 644 |
+
"grad_norm": 0.18498662114143372,
|
| 645 |
+
"learning_rate": 4.48241358831617e-05,
|
| 646 |
+
"entropy": 0.20072167329490184,
|
| 647 |
+
"num_tokens": 27024799.0,
|
| 648 |
+
"mean_token_accuracy": 0.9514979749917984,
|
| 649 |
+
"epoch": 0.5752212389380531,
|
| 650 |
+
"step": 650
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"loss": 0.2368,
|
| 654 |
+
"grad_norm": 0.27973467111587524,
|
| 655 |
+
"learning_rate": 4.3309926530667586e-05,
|
| 656 |
+
"entropy": 0.24167309887707233,
|
| 657 |
+
"num_tokens": 27423001.0,
|
| 658 |
+
"mean_token_accuracy": 0.9416618674993515,
|
| 659 |
+
"epoch": 0.584070796460177,
|
| 660 |
+
"step": 660
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"loss": 0.275,
|
| 664 |
+
"grad_norm": 0.25306880474090576,
|
| 665 |
+
"learning_rate": 4.18019405036366e-05,
|
| 666 |
+
"entropy": 0.2826690062880516,
|
| 667 |
+
"num_tokens": 27822869.0,
|
| 668 |
+
"mean_token_accuracy": 0.9324941843748092,
|
| 669 |
+
"epoch": 0.5929203539823009,
|
| 670 |
+
"step": 670
|
| 671 |
+
},
|
| 672 |
+
{
|
| 673 |
+
"loss": 0.2537,
|
| 674 |
+
"grad_norm": 0.20408447086811066,
|
| 675 |
+
"learning_rate": 4.030158058006262e-05,
|
| 676 |
+
"entropy": 0.25533573105931284,
|
| 677 |
+
"num_tokens": 28246631.0,
|
| 678 |
+
"mean_token_accuracy": 0.9390394508838653,
|
| 679 |
+
"epoch": 0.6017699115044248,
|
| 680 |
+
"step": 680
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"loss": 0.1987,
|
| 684 |
+
"grad_norm": 0.2749539017677307,
|
| 685 |
+
"learning_rate": 3.881024244388827e-05,
|
| 686 |
+
"entropy": 0.2049507148563862,
|
| 687 |
+
"num_tokens": 28686952.0,
|
| 688 |
+
"mean_token_accuracy": 0.9508672833442688,
|
| 689 |
+
"epoch": 0.6106194690265486,
|
| 690 |
+
"step": 690
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"loss": 0.2304,
|
| 694 |
+
"grad_norm": 0.1964893341064453,
|
| 695 |
+
"learning_rate": 3.7329313386694065e-05,
|
| 696 |
+
"entropy": 0.23374196365475655,
|
| 697 |
+
"num_tokens": 29090626.0,
|
| 698 |
+
"mean_token_accuracy": 0.9433914422988892,
|
| 699 |
+
"epoch": 0.6194690265486725,
|
| 700 |
+
"step": 700
|
| 701 |
+
},
|
| 702 |
+
{
|
| 703 |
+
"loss": 0.205,
|
| 704 |
+
"grad_norm": 0.18834583461284637,
|
| 705 |
+
"learning_rate": 3.586017101719432e-05,
|
| 706 |
+
"entropy": 0.21235244944691659,
|
| 707 |
+
"num_tokens": 29510980.0,
|
| 708 |
+
"mean_token_accuracy": 0.9498258531093597,
|
| 709 |
+
"epoch": 0.6283185840707964,
|
| 710 |
+
"step": 710
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"loss": 0.2627,
|
| 714 |
+
"grad_norm": 0.23034313321113586,
|
| 715 |
+
"learning_rate": 3.440418197974039e-05,
|
| 716 |
+
"entropy": 0.26501770280301573,
|
| 717 |
+
"num_tokens": 29922686.0,
|
| 718 |
+
"mean_token_accuracy": 0.9365832090377808,
|
| 719 |
+
"epoch": 0.6371681415929203,
|
| 720 |
+
"step": 720
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"loss": 0.2408,
|
| 724 |
+
"grad_norm": 0.33568480610847473,
|
| 725 |
+
"learning_rate": 3.2962700683023376e-05,
|
| 726 |
+
"entropy": 0.2463029097765684,
|
| 727 |
+
"num_tokens": 30338628.0,
|
| 728 |
+
"mean_token_accuracy": 0.940551894903183,
|
| 729 |
+
"epoch": 0.6460176991150443,
|
| 730 |
+
"step": 730
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"loss": 0.2326,
|
| 734 |
+
"grad_norm": 0.22903338074684143,
|
| 735 |
+
"learning_rate": 3.153706804015873e-05,
|
| 736 |
+
"entropy": 0.23602683916687967,
|
| 737 |
+
"num_tokens": 30751615.0,
|
| 738 |
+
"mean_token_accuracy": 0.9433848887681962,
|
| 739 |
+
"epoch": 0.6548672566371682,
|
| 740 |
+
"step": 740
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"loss": 0.283,
|
| 744 |
+
"grad_norm": 0.24895504117012024,
|
| 745 |
+
"learning_rate": 3.0128610221325022e-05,
|
| 746 |
+
"entropy": 0.28722366876900196,
|
| 747 |
+
"num_tokens": 31149202.0,
|
| 748 |
+
"mean_token_accuracy": 0.9306167840957642,
|
| 749 |
+
"epoch": 0.6637168141592921,
|
| 750 |
+
"step": 750
|
| 751 |
+
},
|
| 752 |
+
{
|
| 753 |
+
"loss": 0.226,
|
| 754 |
+
"grad_norm": 0.20008179545402527,
|
| 755 |
+
"learning_rate": 2.873863742011696e-05,
|
| 756 |
+
"entropy": 0.23511841669678687,
|
| 757 |
+
"num_tokens": 31550325.0,
|
| 758 |
+
"mean_token_accuracy": 0.9445117026567459,
|
| 759 |
+
"epoch": 0.672566371681416,
|
| 760 |
+
"step": 760
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"loss": 0.183,
|
| 764 |
+
"grad_norm": 0.1967935413122177,
|
| 765 |
+
"learning_rate": 2.7368442634760438e-05,
|
| 766 |
+
"entropy": 0.18726294599473475,
|
| 767 |
+
"num_tokens": 31986135.0,
|
| 768 |
+
"mean_token_accuracy": 0.9549686163663864,
|
| 769 |
+
"epoch": 0.6814159292035398,
|
| 770 |
+
"step": 770
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"loss": 0.2182,
|
| 774 |
+
"grad_norm": 0.18984965980052948,
|
| 775 |
+
"learning_rate": 2.6019300465323177e-05,
|
| 776 |
+
"entropy": 0.22221928611397743,
|
| 777 |
+
"num_tokens": 32415282.0,
|
| 778 |
+
"mean_token_accuracy": 0.9468269884586334,
|
| 779 |
+
"epoch": 0.6902654867256637,
|
| 780 |
+
"step": 780
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"loss": 0.2005,
|
| 784 |
+
"grad_norm": 0.22404947876930237,
|
| 785 |
+
"learning_rate": 2.4692465928040043e-05,
|
| 786 |
+
"entropy": 0.20175706930458545,
|
| 787 |
+
"num_tokens": 32833217.0,
|
| 788 |
+
"mean_token_accuracy": 0.951060900092125,
|
| 789 |
+
"epoch": 0.6991150442477876,
|
| 790 |
+
"step": 790
|
| 791 |
+
},
|
| 792 |
+
{
|
| 793 |
+
"loss": 0.21,
|
| 794 |
+
"grad_norm": 0.16944018006324768,
|
| 795 |
+
"learning_rate": 2.3389173287855825e-05,
|
| 796 |
+
"entropy": 0.21427074801176788,
|
| 797 |
+
"num_tokens": 33241388.0,
|
| 798 |
+
"mean_token_accuracy": 0.9489449441432953,
|
| 799 |
+
"epoch": 0.7079646017699115,
|
| 800 |
+
"step": 800
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"loss": 0.2413,
|
| 804 |
+
"grad_norm": 0.2791028916835785,
|
| 805 |
+
"learning_rate": 2.2110634910271554e-05,
|
| 806 |
+
"entropy": 0.24735822584480047,
|
| 807 |
+
"num_tokens": 33638748.0,
|
| 808 |
+
"mean_token_accuracy": 0.9416136890649796,
|
| 809 |
+
"epoch": 0.7168141592920354,
|
| 810 |
+
"step": 810
|
| 811 |
+
},
|
| 812 |
+
{
|
| 813 |
+
"loss": 0.2428,
|
| 814 |
+
"grad_norm": 0.17646969854831696,
|
| 815 |
+
"learning_rate": 2.0858040133562383e-05,
|
| 816 |
+
"entropy": 0.24746169298887252,
|
| 817 |
+
"num_tokens": 34065988.0,
|
| 818 |
+
"mean_token_accuracy": 0.9409119069576264,
|
| 819 |
+
"epoch": 0.7256637168141593,
|
| 820 |
+
"step": 820
|
| 821 |
+
},
|
| 822 |
+
{
|
| 823 |
+
"loss": 0.2431,
|
| 824 |
+
"grad_norm": 0.25427207350730896,
|
| 825 |
+
"learning_rate": 1.963255416241626e-05,
|
| 826 |
+
"entropy": 0.24203080534934998,
|
| 827 |
+
"num_tokens": 34488707.0,
|
| 828 |
+
"mean_token_accuracy": 0.9407094061374665,
|
| 829 |
+
"epoch": 0.7345132743362832,
|
| 830 |
+
"step": 830
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"loss": 0.2239,
|
| 834 |
+
"grad_norm": 0.21476761996746063,
|
| 835 |
+
"learning_rate": 1.843531698402222e-05,
|
| 836 |
+
"entropy": 0.22717048823833466,
|
| 837 |
+
"num_tokens": 34909244.0,
|
| 838 |
+
"mean_token_accuracy": 0.9451982468366623,
|
| 839 |
+
"epoch": 0.7433628318584071,
|
| 840 |
+
"step": 840
|
| 841 |
+
},
|
| 842 |
+
{
|
| 843 |
+
"loss": 0.2208,
|
| 844 |
+
"grad_norm": 0.22185730934143066,
|
| 845 |
+
"learning_rate": 1.7267442307617082e-05,
|
| 846 |
+
"entropy": 0.22914122715592383,
|
| 847 |
+
"num_tokens": 35349557.0,
|
| 848 |
+
"mean_token_accuracy": 0.9452274709939956,
|
| 849 |
+
"epoch": 0.7522123893805309,
|
| 850 |
+
"step": 850
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"loss": 0.2294,
|
| 854 |
+
"grad_norm": 0.1716049462556839,
|
| 855 |
+
"learning_rate": 1.613001652847658e-05,
|
| 856 |
+
"entropy": 0.23541589826345444,
|
| 857 |
+
"num_tokens": 35775557.0,
|
| 858 |
+
"mean_token_accuracy": 0.9432215124368668,
|
| 859 |
+
"epoch": 0.7610619469026548,
|
| 860 |
+
"step": 860
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"loss": 0.3027,
|
| 864 |
+
"grad_norm": 0.22304639220237732,
|
| 865 |
+
"learning_rate": 1.5024097717314894e-05,
|
| 866 |
+
"entropy": 0.3032771345227957,
|
| 867 |
+
"num_tokens": 36187680.0,
|
| 868 |
+
"mean_token_accuracy": 0.9272681981325149,
|
| 869 |
+
"epoch": 0.7699115044247787,
|
| 870 |
+
"step": 870
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"loss": 0.2042,
|
| 874 |
+
"grad_norm": 0.21043775975704193,
|
| 875 |
+
"learning_rate": 1.3950714636032691e-05,
|
| 876 |
+
"entropy": 0.20667122304439545,
|
| 877 |
+
"num_tokens": 36599241.0,
|
| 878 |
+
"mean_token_accuracy": 0.950884211063385,
|
| 879 |
+
"epoch": 0.7787610619469026,
|
| 880 |
+
"step": 880
|
| 881 |
+
},
|
| 882 |
+
{
|
| 883 |
+
"loss": 0.2206,
|
| 884 |
+
"grad_norm": 0.19236122071743011,
|
| 885 |
+
"learning_rate": 1.2910865780728998e-05,
|
| 886 |
+
"entropy": 0.22606479078531266,
|
| 887 |
+
"num_tokens": 37006428.0,
|
| 888 |
+
"mean_token_accuracy": 0.9452178955078125,
|
| 889 |
+
"epoch": 0.7876106194690266,
|
| 890 |
+
"step": 890
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"loss": 0.1997,
|
| 894 |
+
"grad_norm": 0.23871539533138275,
|
| 895 |
+
"learning_rate": 1.1905518452867475e-05,
|
| 896 |
+
"entropy": 0.20729146413505078,
|
| 897 |
+
"num_tokens": 37403547.0,
|
| 898 |
+
"mean_token_accuracy": 0.9499319463968277,
|
| 899 |
+
"epoch": 0.7964601769911505,
|
| 900 |
+
"step": 900
|
| 901 |
+
},
|
| 902 |
+
{
|
| 903 |
+
"loss": 0.2453,
|
| 904 |
+
"grad_norm": 0.18100206553936005,
|
| 905 |
+
"learning_rate": 1.0935607859460984e-05,
|
| 906 |
+
"entropy": 0.2504092514514923,
|
| 907 |
+
"num_tokens": 37815049.0,
|
| 908 |
+
"mean_token_accuracy": 0.9400516986846924,
|
| 909 |
+
"epoch": 0.8053097345132744,
|
| 910 |
+
"step": 910
|
| 911 |
+
},
|
| 912 |
+
{
|
| 913 |
+
"loss": 0.1926,
|
| 914 |
+
"grad_norm": 0.1908176839351654,
|
| 915 |
+
"learning_rate": 1.0002036243111251e-05,
|
| 916 |
+
"entropy": 0.1982258576899767,
|
| 917 |
+
"num_tokens": 38229740.0,
|
| 918 |
+
"mean_token_accuracy": 0.9525273025035859,
|
| 919 |
+
"epoch": 0.8141592920353983,
|
| 920 |
+
"step": 920
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"loss": 0.2521,
|
| 924 |
+
"grad_norm": 0.1694677472114563,
|
| 925 |
+
"learning_rate": 9.10567204271336e-06,
|
| 926 |
+
"entropy": 0.25635765232145785,
|
| 927 |
+
"num_tokens": 38652542.0,
|
| 928 |
+
"mean_token_accuracy": 0.9399167329072953,
|
| 929 |
+
"epoch": 0.8230088495575221,
|
| 930 |
+
"step": 930
|
| 931 |
+
},
|
| 932 |
+
{
|
| 933 |
+
"loss": 0.2137,
|
| 934 |
+
"grad_norm": 0.27372586727142334,
|
| 935 |
+
"learning_rate": 8.247349085605389e-06,
|
| 936 |
+
"entropy": 0.21369801536202432,
|
| 937 |
+
"num_tokens": 39046941.0,
|
| 938 |
+
"mean_token_accuracy": 0.94815693795681,
|
| 939 |
+
"epoch": 0.831858407079646,
|
| 940 |
+
"step": 940
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"loss": 0.2466,
|
| 944 |
+
"grad_norm": 0.20555748045444489,
|
| 945 |
+
"learning_rate": 7.4278658119149695e-06,
|
| 946 |
+
"entropy": 0.24944488294422626,
|
| 947 |
+
"num_tokens": 39471447.0,
|
| 948 |
+
"mean_token_accuracy": 0.9402138769626618,
|
| 949 |
+
"epoch": 0.8407079646017699,
|
| 950 |
+
"step": 950
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"loss": 0.1957,
|
| 954 |
+
"grad_norm": 0.1944950520992279,
|
| 955 |
+
"learning_rate": 6.647984531824064e-06,
|
| 956 |
+
"entropy": 0.20089373365044594,
|
| 957 |
+
"num_tokens": 39880021.0,
|
| 958 |
+
"mean_token_accuracy": 0.9509478449821472,
|
| 959 |
+
"epoch": 0.8495575221238938,
|
| 960 |
+
"step": 960
|
| 961 |
+
},
|
| 962 |
+
{
|
| 963 |
+
"loss": 0.2061,
|
| 964 |
+
"grad_norm": 0.17296965420246124,
|
| 965 |
+
"learning_rate": 5.908430716443086e-06,
|
| 966 |
+
"entropy": 0.21387975346297025,
|
| 967 |
+
"num_tokens": 40310388.0,
|
| 968 |
+
"mean_token_accuracy": 0.9484301716089248,
|
| 969 |
+
"epoch": 0.8584070796460177,
|
| 970 |
+
"step": 970
|
| 971 |
+
},
|
| 972 |
+
{
|
| 973 |
+
"loss": 0.2323,
|
| 974 |
+
"grad_norm": 0.20349572598934174,
|
| 975 |
+
"learning_rate": 5.20989232295393e-06,
|
| 976 |
+
"entropy": 0.23771359361708164,
|
| 977 |
+
"num_tokens": 40710959.0,
|
| 978 |
+
"mean_token_accuracy": 0.94212906062603,
|
| 979 |
+
"epoch": 0.8672566371681416,
|
| 980 |
+
"step": 980
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"loss": 0.2377,
|
| 984 |
+
"grad_norm": 0.18183237314224243,
|
| 985 |
+
"learning_rate": 4.5530191546496515e-06,
|
| 986 |
+
"entropy": 0.2454454731196165,
|
| 987 |
+
"num_tokens": 41138761.0,
|
| 988 |
+
"mean_token_accuracy": 0.9407603412866592,
|
| 989 |
+
"epoch": 0.8761061946902655,
|
| 990 |
+
"step": 990
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"loss": 0.2071,
|
| 994 |
+
"grad_norm": 0.177731454372406,
|
| 995 |
+
"learning_rate": 3.938422256466185e-06,
|
| 996 |
+
"entropy": 0.21258567087352276,
|
| 997 |
+
"num_tokens": 41549796.0,
|
| 998 |
+
"mean_token_accuracy": 0.948466694355011,
|
| 999 |
+
"epoch": 0.8849557522123894,
|
| 1000 |
+
"step": 1000
|
| 1001 |
+
},
|
| 1002 |
+
{
|
| 1003 |
+
"loss": 0.2984,
|
| 1004 |
+
"grad_norm": 0.18465429544448853,
|
| 1005 |
+
"learning_rate": 3.3666733465682833e-06,
|
| 1006 |
+
"entropy": 0.29959365651011466,
|
| 1007 |
+
"num_tokens": 41982205.0,
|
| 1008 |
+
"mean_token_accuracy": 0.9291581392288208,
|
| 1009 |
+
"epoch": 0.8938053097345132,
|
| 1010 |
+
"step": 1010
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"loss": 0.223,
|
| 1014 |
+
"grad_norm": 0.18058468401432037,
|
| 1015 |
+
"learning_rate": 2.8383042845186004e-06,
|
| 1016 |
+
"entropy": 0.22986317798495293,
|
| 1017 |
+
"num_tokens": 42424950.0,
|
| 1018 |
+
"mean_token_accuracy": 0.9453666418790817,
|
| 1019 |
+
"epoch": 0.9026548672566371,
|
| 1020 |
+
"step": 1020
|
| 1021 |
+
},
|
| 1022 |
+
{
|
| 1023 |
+
"loss": 0.221,
|
| 1024 |
+
"grad_norm": 0.19499893486499786,
|
| 1025 |
+
"learning_rate": 2.3538065765244755e-06,
|
| 1026 |
+
"entropy": 0.22309133298695089,
|
| 1027 |
+
"num_tokens": 42834983.0,
|
| 1028 |
+
"mean_token_accuracy": 0.9468386858701706,
|
| 1029 |
+
"epoch": 0.911504424778761,
|
| 1030 |
+
"step": 1030
|
| 1031 |
+
},
|
| 1032 |
+
{
|
| 1033 |
+
"loss": 0.2156,
|
| 1034 |
+
"grad_norm": 0.19273503124713898,
|
| 1035 |
+
"learning_rate": 1.913630918222792e-06,
|
| 1036 |
+
"entropy": 0.21915269698947668,
|
| 1037 |
+
"num_tokens": 43234797.0,
|
| 1038 |
+
"mean_token_accuracy": 0.9469860643148422,
|
| 1039 |
+
"epoch": 0.9203539823008849,
|
| 1040 |
+
"step": 1040
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"loss": 0.2552,
|
| 1044 |
+
"grad_norm": 0.15608751773834229,
|
| 1045 |
+
"learning_rate": 1.5181867754280931e-06,
|
| 1046 |
+
"entropy": 0.25907820984721186,
|
| 1047 |
+
"num_tokens": 43679426.0,
|
| 1048 |
+
"mean_token_accuracy": 0.9379706591367721,
|
| 1049 |
+
"epoch": 0.9292035398230089,
|
| 1050 |
+
"step": 1050
|
| 1051 |
+
},
|
| 1052 |
+
{
|
| 1053 |
+
"loss": 0.2416,
|
| 1054 |
+
"grad_norm": 0.1610129177570343,
|
| 1055 |
+
"learning_rate": 1.1678420032341153e-06,
|
| 1056 |
+
"entropy": 0.24386902302503585,
|
| 1057 |
+
"num_tokens": 44112256.0,
|
| 1058 |
+
"mean_token_accuracy": 0.9416323363780975,
|
| 1059 |
+
"epoch": 0.9380530973451328,
|
| 1060 |
+
"step": 1060
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"loss": 0.2071,
|
| 1064 |
+
"grad_norm": 0.16213390231132507,
|
| 1065 |
+
"learning_rate": 8.629225038229049e-07,
|
| 1066 |
+
"entropy": 0.20874513685703278,
|
| 1067 |
+
"num_tokens": 44570422.0,
|
| 1068 |
+
"mean_token_accuracy": 0.9496494054794311,
|
| 1069 |
+
"epoch": 0.9469026548672567,
|
| 1070 |
+
"step": 1070
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"loss": 0.1747,
|
| 1074 |
+
"grad_norm": 0.18272142112255096,
|
| 1075 |
+
"learning_rate": 6.037119232999266e-07,
|
| 1076 |
+
"entropy": 0.17933553121984006,
|
| 1077 |
+
"num_tokens": 44972484.0,
|
| 1078 |
+
"mean_token_accuracy": 0.9556117475032806,
|
| 1079 |
+
"epoch": 0.9557522123893806,
|
| 1080 |
+
"step": 1080
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"loss": 0.2177,
|
| 1084 |
+
"grad_norm": 0.21331429481506348,
|
| 1085 |
+
"learning_rate": 3.904513878371818e-07,
|
| 1086 |
+
"entropy": 0.2203224040567875,
|
| 1087 |
+
"num_tokens": 45399966.0,
|
| 1088 |
+
"mean_token_accuracy": 0.9465226858854294,
|
| 1089 |
+
"epoch": 0.9646017699115044,
|
| 1090 |
+
"step": 1090
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"loss": 0.2361,
|
| 1094 |
+
"grad_norm": 0.19762705266475677,
|
| 1095 |
+
"learning_rate": 2.233392793697442e-07,
|
| 1096 |
+
"entropy": 0.23971713967621328,
|
| 1097 |
+
"num_tokens": 45841913.0,
|
| 1098 |
+
"mean_token_accuracy": 0.9428269326686859,
|
| 1099 |
+
"epoch": 0.9734513274336283,
|
| 1100 |
+
"step": 1100
|
| 1101 |
+
},
|
| 1102 |
+
{
|
| 1103 |
+
"loss": 0.254,
|
| 1104 |
+
"grad_norm": 0.21293021738529205,
|
| 1105 |
+
"learning_rate": 1.0253105105438865e-07,
|
| 1106 |
+
"entropy": 0.25558883510529995,
|
| 1107 |
+
"num_tokens": 46272798.0,
|
| 1108 |
+
"mean_token_accuracy": 0.9393420517444611,
|
| 1109 |
+
"epoch": 0.9823008849557522,
|
| 1110 |
+
"step": 1110
|
| 1111 |
+
},
|
| 1112 |
+
{
|
| 1113 |
+
"loss": 0.2335,
|
| 1114 |
+
"grad_norm": 0.1932452768087387,
|
| 1115 |
+
"learning_rate": 2.8139082661954307e-08,
|
| 1116 |
+
"entropy": 0.24025008846074342,
|
| 1117 |
+
"num_tokens": 46703041.0,
|
| 1118 |
+
"mean_token_accuracy": 0.9413352489471436,
|
| 1119 |
+
"epoch": 0.9911504424778761,
|
| 1120 |
+
"step": 1120
|
| 1121 |
+
},
|
| 1122 |
+
{
|
| 1123 |
+
"loss": 0.2372,
|
| 1124 |
+
"grad_norm": 0.1784687340259552,
|
| 1125 |
+
"learning_rate": 2.32576038022847e-10,
|
| 1126 |
+
"entropy": 0.23823871053755283,
|
| 1127 |
+
"num_tokens": 47110855.0,
|
| 1128 |
+
"mean_token_accuracy": 0.943606698513031,
|
| 1129 |
+
"epoch": 1.0,
|
| 1130 |
+
"step": 1130
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"train_runtime": 18643.3649,
|
| 1134 |
+
"train_samples_per_second": 1.94,
|
| 1135 |
+
"train_steps_per_second": 0.061,
|
| 1136 |
+
"total_flos": 2.0704071601101865e+19,
|
| 1137 |
+
"train_loss": 0.33413480024422165,
|
| 1138 |
+
"epoch": 1.0,
|
| 1139 |
+
"step": 1130
|
| 1140 |
+
}
|
| 1141 |
+
]
|